
from ultralytics import YOLO
from PIL import Image
import random
import os


model = YOLO("yolov8x.pt")
model.predict("cat_dog.jpg", save=True, project=".", name="yolov8x")
pass

image 1/1 /home/blazej/Desktop/obrazy/lab4/cat_dog.jpg: 448x640 1 cat, 1 dog, 100.7ms
Speed: 1.3ms preprocess, 100.7ms inference, 1.3ms postprocess per image at shape (1, 3, 448, 640)
Results saved to yolov8x


Image.open("yolov8x/cat_dog.jpg")


%pip install boto3 botocore
!curl https://raw.githubusercontent.com/openimages/dataset/master/downloader.py -o downloader.py


# Get boxable class names
# !curl https://storage.googleapis.com/openimages/v7/oidv7-class-descriptions-boxable.csv -o metadata/class_names.csv
!cat metadata/class_names.csv | grep Food

/m/02wbm,Food
/m/03y6mg,Food processor


CATEGORY = "/m/02wbm"


os.makedirs("metadata", exist_ok=True)


!curl https://storage.googleapis.com/openimages/v6/oidv6-train-annotations-bbox.csv -o metadata/train_boxes.csv
!curl https://storage.googleapis.com/openimages/v5/validation-annotations-bbox.csv -o metadata/validation_boxes.csv
!curl https://storage.googleapis.com/openimages/v5/test-annotations-bbox.csv -o metadata/test_boxes.csv


DATASET_PATH = "datasets/food"


os.makedirs(DATASET_PATH + "/train/images", exist_ok=True)
os.makedirs(DATASET_PATH + "/train/labels", exist_ok=True)
os.makedirs(DATASET_PATH + "/validation/images", exist_ok=True)
os.makedirs(DATASET_PATH + "/validation/labels", exist_ok=True)


def get_metadata(set: str):
    assert set in ["train", "validation", "test"]

    metadata = []

    with open(f"metadata/{set}_boxes.csv", "r") as file:
        lines = file.readlines()[1:]
        for line in lines:
            fields = line.strip().split(",")
            if len(fields) > 1 and fields[2] == CATEGORY:
                metadata.append(fields)
    
    return metadata


def prepare_labels_and_ids(set: str, metadata: list[list[str]], size: int):
    assert set in ["train", "validation", "test"]
    labels: dict[str, list[str]] = {}

    for fields in metadata:
        xmin, xmax, ymin, ymax = map(float, fields[4:8])
        image_id = fields[0]

        x = (xmin + xmax) / 2
        y = (ymin + ymax) / 2
        w = xmax - xmin
        h = ymax - ymin

        row = f"0 {x} {y} {w} {h}\n"

        if image_id in labels:
            labels[image_id].append(row)
        else:
            labels[image_id] = [row]

    selected_images = list(labels.items())[:size]

    for id, rows in selected_images:
        with open(f"{DATASET_PATH}/{set}/labels/{id}.txt", "w") as file:
            file.writelines(rows)


    ids = [f"{set}/{id}\n" for id, _ in selected_images]

    with open(f"{DATASET_PATH}/{set}/ids.txt", "w") as file:
        file.writelines(ids)


metadata = get_metadata("train")
prepare_labels_and_ids("train", metadata, 2000)


!python downloader.py datasets/food/train/ids.txt --download_folder datasets/food/train/images --num_processes 5


metadata = get_metadata("validation")
prepare_labels_and_ids("validation", metadata, 200)


!python downloader.py datasets/food/validation/ids.txt --download_folder datasets/food/validation/images --num_processes 5


config = '''\
train: ./food/train
val: ./food/validation
nc: 1 
names: ['food']
'''

with open("food.yaml", "w") as f:
    f.write(config)


from ultralytics import YOLO
import torch
torch.backends.cudnn.enabled = False

# Load a model
model = YOLO("yolov8n.pt")  # load a pretrained model. change this to the model you want to use, n, s, l etc. (only ending)

# Train the model
results = model.train(data=f"food.yaml", epochs=10)

Ultralytics YOLOv8.0.212 🚀 Python-3.9.18 torch-2.1.0 CUDA:0 (NVIDIA GeForce GTX 1650, 3906MiB)
engine/trainer: task=detect, mode=train, model=yolov8n.pt, data=food.yaml, epochs=10, patience=50, batch=16, imgsz=640, save=True, save_period=-1, cache=False, device=None, workers=8, project=None, name=train2, exist_ok=False, pretrained=True, optimizer=auto, verbose=True, seed=0, deterministic=True, single_cls=False, rect=False, cos_lr=False, close_mosaic=10, resume=False, amp=True, fraction=1.0, profile=False, freeze=None, overlap_mask=True, mask_ratio=4, dropout=0.0, val=True, split=val, save_json=False, save_hybrid=False, conf=None, iou=0.7, max_det=300, half=False, dnn=False, plots=True, source=None, show=False, save_txt=False, save_conf=False, save_crop=False, show_labels=True, show_conf=True, vid_stride=1, stream_buffer=False, line_width=None, visualize=False, augment=False, agnostic_nms=False, classes=None, retina_masks=False, boxes=True, format=torchscript, keras=False, optimize=False, int8=False, dynamic=False, simplify=False, opset=None, workspace=4, nms=False, lr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=7.5, cls=0.5, dfl=1.5, pose=12.0, kobj=1.0, label_smoothing=0.0, nbs=64, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, mosaic=1.0, mixup=0.0, copy_paste=0.0, cfg=None, tracker=botsort.yaml, save_dir=runs/detect/train2
Overriding model.yaml nc=80 with nc=1

                   from  n    params  module                                       arguments                     
  0                  -1  1       464  ultralytics.nn.modules.conv.Conv             [3, 16, 3, 2]                 
  1                  -1  1      4672  ultralytics.nn.modules.conv.Conv             [16, 32, 3, 2]                
  2                  -1  1      7360  ultralytics.nn.modules.block.C2f             [32, 32, 1, True]             
  3                  -1  1     18560  ultralytics.nn.modules.conv.Conv             [32, 64, 3, 2]                
  4                  -1  2     49664  ultralytics.nn.modules.block.C2f             [64, 64, 2, True]             
  5                  -1  1     73984  ultralytics.nn.modules.conv.Conv             [64, 128, 3, 2]               
  6                  -1  2    197632  ultralytics.nn.modules.block.C2f             [128, 128, 2, True]           
  7                  -1  1    295424  ultralytics.nn.modules.conv.Conv             [128, 256, 3, 2]              
  8                  -1  1    460288  ultralytics.nn.modules.block.C2f             [256, 256, 1, True]           
  9                  -1  1    164608  ultralytics.nn.modules.block.SPPF            [256, 256, 5]                 
 10                  -1  1         0  torch.nn.modules.upsampling.Upsample         [None, 2, 'nearest']          
 11             [-1, 6]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 12                  -1  1    148224  ultralytics.nn.modules.block.C2f             [384, 128, 1]                 
 13                  -1  1         0  torch.nn.modules.upsampling.Upsample         [None, 2, 'nearest']          
 14             [-1, 4]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 15                  -1  1     37248  ultralytics.nn.modules.block.C2f             [192, 64, 1]                  
 16                  -1  1     36992  ultralytics.nn.modules.conv.Conv             [64, 64, 3, 2]                
 17            [-1, 12]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 18                  -1  1    123648  ultralytics.nn.modules.block.C2f             [192, 128, 1]                 
 19                  -1  1    147712  ultralytics.nn.modules.conv.Conv             [128, 128, 3, 2]              
 20             [-1, 9]  1         0  ultralytics.nn.modules.conv.Concat           [1]                           
 21                  -1  1    493056  ultralytics.nn.modules.block.C2f             [384, 256, 1]                 
 22        [15, 18, 21]  1    751507  ultralytics.nn.modules.head.Detect           [1, [64, 128, 256]]           
Model summary: 225 layers, 3011043 parameters, 3011027 gradients, 8.2 GFLOPs

Transferred 319/355 items from pretrained weights
Freezing layer 'model.22.dfl.conv.weight'
AMP: running Automatic Mixed Precision (AMP) checks with YOLOv8n...
AMP: checks passed ✅

train: Scanning /home/blazej/Desktop/obrazy/lab4/datasets/food/train/labels.cache... 2000 images, 0 backgrounds, 0 corrupt: 100%|██████████| 2000/2000 [00:00<?, ?it/s]
val: Scanning /home/blazej/Desktop/obrazy/lab4/datasets/food/validation/labels.cache... 200 images, 0 backgrounds, 0 corrupt: 100%|██████████| 200/200 [00:00<?, ?it/s]

Plotting labels to runs/detect/train2/labels.jpg... 
optimizer: 'optimizer=auto' found, ignoring 'lr0=0.01' and 'momentum=0.937' and determining best 'optimizer', 'lr0' and 'momentum' automatically... 
optimizer: AdamW(lr=0.002, momentum=0.9) with parameter groups 57 weight(decay=0.0), 64 weight(decay=0.0005), 63 bias(decay=0.0)
Image sizes 640 train, 640 val
Using 8 dataloader workers
Logging results to runs/detect/train2
Starting training for 10 epochs...
Closing dataloader mosaic

      Epoch    GPU_mem   box_loss   cls_loss   dfl_loss  Instances       Size

       1/10      2.62G      1.474       2.39      1.543         74        640: 100%|██████████| 125/125 [03:14<00:00,  1.55s/it]
                 Class     Images  Instances      Box(P          R      mAP50  mAP50-95): 100%|██████████| 7/7 [00:07<00:00,  1.01s/it]

                   all        200        324       0.27      0.327      0.183     0.0637

      Epoch    GPU_mem   box_loss   cls_loss   dfl_loss  Instances       Size

       2/10      2.68G      1.635      2.134      1.653         33        640: 100%|██████████| 125/125 [03:21<00:00,  1.61s/it]
                 Class     Images  Instances      Box(P          R      mAP50  mAP50-95): 100%|██████████| 7/7 [00:06<00:00,  1.01it/s]

                   all        200        324      0.152      0.203     0.0805      0.031


from PIL import Image
Image.open("runs/detect/train2/val_batch0_pred.jpg")


import os
import random
from ultralytics import YOLO
import cv2
from matplotlib import pyplot as plt

def blur_food(image_path):
    model = YOLO("runs/detect/train2/weights/last.pt")
    preds = model.predict(image_path)

    boxes = preds[0].boxes

    image = cv2.imread(image_path)

    for box in boxes:
        x, y, w, h = map(int, box.xywh.tolist()[0])
        x = x - w//2
        y = y - h//2
        image[y:y+h, x:x+w] = cv2.GaussianBlur(image[y:y+h, x:x+w], (15, 15),cv2.BORDER_DEFAULT)

    plt.figure(figsize = (15,10))
    plt.imshow(image[:, :, ::-1], interpolation='nearest')


blur_food("datasets/food/validation/images/1681f749e331106f.jpg")

image 1/1 /home/blazej/Desktop/obrazy/lab4/datasets/food/validation/images/1681f749e331106f.jpg: 640x448 3 foods, 9.1ms
Speed: 2.9ms preprocess, 9.1ms inference, 1.5ms postprocess per image at shape (1, 3, 640, 448)


blur_food("datasets/food/validation/images/0550b879165d4392.jpg")

image 1/1 /home/blazej/Desktop/obrazy/lab4/datasets/food/validation/images/0550b879165d4392.jpg: 640x640 17 foods, 8.2ms
Speed: 3.8ms preprocess, 8.2ms inference, 1.6ms postprocess per image at shape (1, 3, 640, 640)


blur_food("datasets/food/validation/images/0b6227bb06345402.jpg")

image 1/1 /home/blazej/Desktop/obrazy/lab4/datasets/food/validation/images/0b6227bb06345402.jpg: 640x448 2 foods, 9.1ms
Speed: 2.5ms preprocess, 9.1ms inference, 1.5ms postprocess per image at shape (1, 3, 640, 448)

Lab 4 - YOLO¶

1. Using retrained model on supported categories¶

2. Create custom dataset with images of food¶